Anyone can explain why the ranking and p-value changed like this?

[Samuel Li]

Hi all,

I run epacts using the following code with 2 vcfs, one with filter base on CADD score and the other didn't

epacts group \
--vcf Rare-variant-0.001-anno.vcf.gz \
--ped $my.ped \
--max-maf 0.05 --pheno DISEASE --test skat --skat-o --skat-adj --out EPACTS/SKATO-adj/2022-6-17-EPACTS-SKATO-pLOF-missense \
--run 16 --cov Age --cov sex --cov PC1 --cov PC2 --cov PC3 --groupf Rare-variant-0.001-anno-pLOF-missense.group

The only 2 changes are the vcfs and the group file as the vcfs are either CADD filtered or not filtered.

And when I get the result (attached), I was confused as SLC17A2 have the same number of variants and faction with rare variant for both scenarios, yet the P-values are different. How is this possible? Any advice would be appreciate

Thank you very much

Samuel Li